Evaluating the Quality of Postpartum Hemorrhage Nursing Care Plans Generated by Artificial Intelligence Models.

Abstract

With the rapidly advancing technological landscape of health care, evaluating the potential use of artificial intelligence (AI) models to prepare nursing care plans is of great importance. The purpose of this study was to evaluate the quality of nursing care plans created by AI for the management of postpartum hemorrhage (PPH). This cross-sectional exploratory study involved creating a scenario for an imaginary patient with PPH. Information was put into 3 AI platforms (GPT-4, LaMDA, Med-PaLM) on consecutive days without prior conversation. Care plans were evaluated using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) scale. Med-PaLM exhibited superior quality in developing the care plan compared with LaMDA ( Z = 4.354; P = .000) and GPT-4 ( Z = 3.126; P = .029). Our findings suggest that despite the strong performance of Med-PaLM, AI, in its current state, is unsuitable for use with real patients.